The Dual Tropism of Noroviruses

Noroviruses are highly prevalent enteric RNA viruses. Human noroviruses (HuNoVs) cause significant morbidity, mortality, and economic losses worldwide.

ABSTRACT

Noroviruses are highly prevalent enteric RNA viruses. Human noroviruses (HuNoVs) cause significant morbidity, mortality, and economic losses worldwide. Infections also occur in other mammalian species, including mice. Despite the discovery of the first norovirus in 1972, the viral tropism has long remained an enigma. A long-held assumption was that these viruses infect intestinal epithelial cells. Recent data support a more complex cell tropism of epithelial and nonepithelial cell types.

INTRODUCTION

Human noroviruses (HuNoVs) cause nearly one-fifth of all cases of diarrhea worldwide, with an estimated global price tag of $60 billion (1, 2). These viruses are estimated to cause ∼200,000 deaths in children under the age of 5 in resource-limited settings (3), and they cause significant morbidity in developed nations, including an estimated 20 million infections each year in the United States alone (4). No specific vaccine or antiviral has been approved to date, and efforts to develop treatment or prevention strategies are hampered by our limited understanding of fundamental features of norovirus biology. The last several decades have seen many exciting advances in the norovirus field. These include the cloning of the Norwalk virus genome (5), the discovery of murine norovirus (MNV) and establishment of MNV cell culture and mouse models (6, 7), and the development of two cell culture systems and multiple animal models for HuNoVs (8–13). However, robust replication models that enable multigenerational passaging of HuNoVs are still lacking (14). A greater understanding of the cell type(s) infected by HuNoVs (i.e., viral tropism) may promote the development of improved cell culture systems, enabling the generation of cell culture-derived virus stocks. This in turn would further advance both basic and translational norovirus research.

Elucidating norovirus tropism has been a focus of research effort since the discovery of the first norovirus, Norwalk virus, in 1972 (15). Given the intestinal symptoms of diarrhea, one focus was on identifying intestinal epithelial cell lines that supported HuNoV replication. The many unsuccessful attempts to identify HuNoV-susceptible intestinal epithelial cell types and susceptible cell types in other established cell lines by two laboratories were summarized in a publication in 2004 to guide subsequent studies (16). The discovery of murine norovirus 1 (MNV-1) in immunocompromised mice in 2003 (6) laid the foundation for a paradigm shift in our understanding of norovirus tropism. The identification of MNV-infected cells with macrophage- and dendritic cell-like morphology in vivo led to the identification of multiple murine macrophage and dendritic cell lines susceptible to MNV-1 in vitro (7). This finding demonstrated that immune cells are susceptible to a norovirus, giving the first indication that noroviruses may have a broader or different tropism than initially anticipated. Subsequent studies with MNV have revealed an intriguing, multifaceted picture of viral tropism.

All MNV strains investigated to date use carbohydrates as attachment receptors, but the glycan receptor specificity varies among virus strains (Table 1) (17–19). MNV-1 and the persistent strain MNV-S99 (20) bind to terminal sialic acids on gangliosides and N- and O-linked glycoproteins, while MNV-CR3 binds to N-linked glycoproteins (18, 19). Following attachment to the cell surface, at least three strains, MNV-1 (specifically its plaque-purified isolate CW3), MNV-CR6, and MNV-S7, use the CD300 family member CD300lf as the viral receptor (Table 1) (21–23). This family of proteins has a single immunoglobulin variable-like (IgV) extracellular domain that is involved in the regulation of immune responses (24, 25). CD300lf is expressed on immune cells and binds phosphatidylserine (PS) to regulate phagocytosis of apoptotic cells and elicit a range of other immune-inhibitory functions. Deletion of CD300lf from cells renders them nonsusceptible to MNV-1 and MNV-CR6, while expression of the molecule converts nonsusceptible cells into susceptible cells (21, 22). CD300ld is a paralog of CD300lf that shares the N-terminal portion of the protein with CD300lf, and it can also confer MNV susceptibility to nonsusceptible cells (22). Thus, the expression of the proteinaceous receptors CD300lf and CD300ld determines the cellular tropism for MNV.

TABLE 1.

Summary of features of select murine norovirus strains

Strain	Receptor(s) used			Infection type	In vivo tropisma
	Attachment receptor(s)		Entry receptor CD300lf
	Sialic-acid linked ganglioside	Glycoproteins
MNV-1b	+	N and O linked	+	Acute	MΦ, DCs, B cells, T cells (acute phase); radiation-resistant cell?
MNV-S99	+	N and O linked	NDc	Persistent	ND
MNV-CR3	−	N linked	ND	Persistent	ND
MNV-CR6	ND	ND	+	Persistent	Tuft cell (persistent phase)
MNV-S7	ND	ND	+	ND	ND
MNV-3	ND	ND	ND	Persistent	ND

^aMΦ, macrophages; DCs, dendritic cells.

^bPlaque isolate MNV-1.CW3.

^cND, not determined.

In vivo, MNV strains have two main infection phenotypes (Table 1). MNV-1 causes an acute infection, and infectious virus is cleared within 1 week (26). Infection is initiated in the distal part of the small intestine and spreads from there to other sites in the small and large intestine, mesenteric lymph nodes, and spleen (27). Replication at early times postinfection (i.e., 24 h) occurs in gut-associated lymphoid tissues (e.g., Peyer's patches), specifically in macrophages, dendritic cells, B cells, and T cells (28). Each of these immune cell types expresses the viral receptor CD300lf (28). Shedding of newly replicated, infectious virions is limited to a very narrow time window between 12 and 24 h postinfection (27). The other phenotype is characterized by a persistent infection, with infectious virus shedding continuing for at least 1 month (26). Most strains fall into this group. Three commonly used persistent strains are MNV-3, MNV-CR3 (also called CR3), and MNV-CR6 (also called CR6). MNV-CR3 initiates infection in the cecum and then spreads to the small intestine, colon, and mesenteric lymph nodes (27). Both the acute MNV-1 strain and the persistent MNV-CR3 strain overcome the intestinal epithelial barrier by hijacking microfold (M) cells in the follicle-associated epithelium overlying Peyer's patches (27, 29, 30). Whether MNV-CR6 and other murine or human norovirus strains follow a similar infection pattern and/or are dependent on M cells for intestinal uptake has not been investigated.

Deletion of CD300lf from mice renders them resistant to MNV-1 and MNV-CR6 infection (21, 31). Studies using bone marrow transplants between wild-type (wt) and CD300lf-deficient mice aimed at identifying the infected cell type revealed that during the persistent phase of the infection (i.e., 21 days postinfection), MNV-CR6 required wt radiation-resistant cells (e.g., epithelial cells) (31). In contrast, detection of the MNV-1 genome at 7 days postinfection is dependent on both radiation-resistant and -nonresistant (e.g., immune cells) cells. Subsequent studies of MNV-CR6 persistence identified a specialized intestinal epithelial cell type, called tuft cell, as the source of infectious virus (31). Tuft cells are a rare type of epithelial cell that express certain immune cell genes, including CD300lf (31). Remarkably, only ∼1.4% of tuft cells in the intestine, or ∼100 cells per mouse, are persistently infected by MNV-CR6 (31, 32). The type 2 cytokines interleukin-25 (IL-25) and IL-4 induce tuft cell proliferation and, thus, promote MNV persistence. These data suggest tuft cells are an immune-privileged cell type in the intestine. This is consistent with previous work demonstrating that gut-resident T cells against MNV-CR6 are functional but ignorant of viral replication during the persistent phase of infection (33). More information is needed to determine the mechanism by which maintenance of a pool of infected tuft cells occurs in an environment of intestinal villi that are continuously renewed. The ability of MNV-CR6 to persist maps to the nonstructural protein NS1, which blocks interferon-lambda (IFN-λ) responses (34–37). Thus, in order for MNV-CR6 to persistently infect tuft cells, it must counteract IFN-λ. However, the mechanism of IFN-λ evasion remains unknown, and studies to date have not identified the cell type(s) that support MNV-CR6 replication during acute infection. Likewise, whether MNV-1 or other strains of MNV can infect tuft cells or another radiation-resistant cell type remains to be determined. Collectively, these data suggest that CD300lf-expressing immune cells in the gastrointestinal lymphoid tissues are the primary cell targets during the acute phase of MNV infection, while persistent MNV hides in CD300lf-expressing tuft cells (Fig. 1).

FIG 1.

The dual tropism of noroviruses for intestinal epithelial and immune cells. As detailed in the text, the acute strain of murine norovirus (MNV), MNV-1, infects macrophages, dendritic cells, and T and B cells in vitro and in gastrointestinal lymphoid-associated tissue, including the Peyer's patch. The persistent strain of MNV, MNV-CR6, establishes persistence in tuft cells. HuNoVs infect intestinal epithelial cells (IEC), macrophages, dendritic cells, and T cells in immunocompromised hosts in vivo and IEC and B cells in vitro. (Adapted from reference 29 with permission.).

HuNoVs historically have been recalcitrant to growth in cell culture. Similar to MNV, HuNoVs attach to host cell glycans, including histo-blood group antigens (HBGAs), which are expressed on the intestinal epithelium, among other tissues (17, 38). However, a putative proteinaceous receptor, which might also mediate cellular tropism, has not been identified. Hence, attempts to culture HuNoVs in intestinal epithelial lines or blood-derived macrophages and dendritic cells have not yielded consistent results (16, 39, 40). It was only recently that infection studies of (i) differentiated human intestinal organoid cultures enriched in mature enterocytes (i.e., absorptive columnar intestinal epithelial cells) and (ii) a subset of established human B cell lines (e.g., BJAB cells) showed evidence of infection (8, 10, 41). Studies in organoids demonstrated the dependence of some HuNoV strains on fucosyltransferase 2 (FUT2) and on bile acids for infection (10). Infection in B cells was improved in the presence of H-type HBGA glycans, either free or associated with bacteria (8). The dual tropism of HuNoV for intestinal epithelial and immune cells in vitro is shared with MNV (28, 31). Whether this dual cellular tropism also extends to infections in vivo has not been addressed fully. Data from animal models support a dual tropism (42). Evidence for infection of the following cell types in animal models has been obtained: dendritic cells and B cells in chimpanzees (11), macrophages in immunocompromised mice (9), enterocytes in gnotobiotic pigs (13), and macrophages, dendritic cells, and lymphocytes in miniature piglets (43). More importantly, histologic analysis of HuNoV-infected immunocompromised patients further supports a dual tropism (44). HuNoV antigen is detected in enterocytes near the villus tips, as well as CD3-, CD68-, or DC-SIGN-positive cells, immune cell markers for T cells, macrophages, and dendritic cells. Whether the same cell types are infected in immunocompetent patients is unresolved at present. Nevertheless, taken together, the data indicate that HuNoVs infect intestinal epithelial cells and several immune cell subsets (macrophages, dendritic cells, T cells, and B cells) in the intestine (Fig. 1).

Although the studies outlined above indicate that human and murine noroviruses share a dual tropism for intestinal epithelial and immune cells, many open questions remain, including those listed below, that will require a variety of approaches and types of expertise to advance our collective understanding of the subject. For example, work on MNV has identified virus strain-specific differences in viral tropism, including persistence in tuft cells. HuNoVs exhibit greater genetic diversity than MNV (26, 45), and thus, strain-specific differences in HuNoVs may lead to different patterns of viral tropism. Therefore, it will be of interest to identify infected cell types for different HuNoV genotypes and strains and to determine whether infection of tuft cells is shared between human and mouse noroviruses. In addition, noroviruses infect many different mammalian species (46). While intestinal epithelial cells and macrophages were identified as infected cell types for bovine norovirus in vivo (47), the tropism of other noroviruses remains undefined. Such in vivo studies may inform in vitro studies for developing cell culture models for non-murine and non-human noroviruses and for improving the systems available for HuNoVs or developing new ones. Advances in organoid technology, including recent efforts to integrate intestinal organoids with immune cells (48) or with an enteric nervous system (49), to more faithfully recapitulate the multiple cell types present in the intestine, will undoubtedly contribute to our increasing understanding of norovirus biology. Thus, the future promises to yield new and exciting findings in the norovirus field that will ultimately result in strategies to effectively control and limit norovirus infections.

OUTSTANDING QUESTIONS

• What is the HuNoV tropism in immunocompetent hosts in vivo?

• Are tuft cells infected by human norovirus?

• Is there a temporal switch in cell types infected by the persistent MNV strains, such that, for example, immune cells are infected first, while tuft cells are infected later?

• Which epithelial and immune cell types are infected by other animal noroviruses?

• What is the contribution of the different cell types to pathogenesis and transmission?

• What is the functional norovirus receptor(s) on the various cell types?

• How does the genetic diversity of noroviruses influence cell and tissue tropism?

• Do glycans and bacteria affect the tropism of noroviruses? If yes, how?